The Internet of Things, which includes environmental sensor networks, industrial monitoring systems, home automation, and wearables, has brought with it an ever-increasing demand for computational power in smart mobile devices. Research in the International Journal of Information and Communication Technology, discusses how mobile edge computing might meet these demands.
Yujie Li, Yaoyao Xu, and Fangfang Cao of the Nanchang Institute of Science and Technology, and Xiang He of Jiangxi Flight University also in Nanchang, China explain how cloud computing served as the backbone for data processing in mobile devices. However, these systems have their shortcomings, such as the requisite connectivity, the likely bottlenecks and delays and the ever-present security and privacy risks associated with having one’s data on a third party system.
Mobile edge computing (MEC) could address many of these issues by decentralizing computation so that instead of a system being reliant on remote data centres and servers, the computational tasks are carried out on edge servers that are closer to the user. This reduces delays, latency, as well as potentially easing the load on the main servers, allowing faster processing of data to take place. However, even this approach raises problems of energy consumption as the mobile devices offload tasks to nearby servers.
The new work from Li and colleagues offers a way to tackle this issue by proposing an innovative resource scheduling model designed to optimize energy use in MEC systems. The new model assigns tasks in a non-random way that takes into consideration the available computational resources of the edge servers and their energy consumption. To enhance this process, the researchers used a sophisticated optimization technique known as enhanced particle swarm optimization. This algorithm gives quick convergence of processes to an optimal solution selected from a broad range of putative solutions to the given problem or task. It is thus more effective in managing energy usage across the entire system – on the mobile devices and on the edge servers.
The researchers’ test results are promising. When applied to a setup with ten smart mobile devices, the model was able to reduce energy consumption by up to 55% compared to other approaches. These savings are achieved through smarter scheduling of computational tasks, ensuring that both the energy needs of the mobile devices and the power costs of the edge servers are balanced.
Li, Y., Xu, Y., Cao, F. and He, X. (2024) ‘A meta-heuristic optimisation algorithm based method for scheduling edge computing resources’, Int. J. Information and Communication Technology, Vol. 25, No. 9, pp.88–103.